In the past few years there has been a change in reservoir modelling from using 2D maps to a 3D workflow involving detailed shared earth models, upscaled for reservoir simulation. This change has brought an improvement in the quality of simulation models through incorporating more geological features of reservoirs. However, in many cases, this process is slow and impractical. Teams find they cannot update the complex models frequently enough, or carry uncertainty, because current technology can not cope. Our task now is to take the shared earth model forward and realise its potential.
This paper introduces a new approach to reservoir modelling: DDRM, which is aimed at meeting the challenges of the new business environment. It involves creating and using reservoir models which are just as complex as they need to be, and bringing uncertainty into predictions more thoroughly than before. This requires advanced reservoir modelling technology, with new techniques such as: scale-independent non-grid-based design, model-detail realisation on the fly, merging static and dynamic elements, and uncertain prediction from uncertain models.
The new approach is illustrated by a case study in probabilistic ranking of well-locations on an oilfield under development. The business decision is where to put the next wells, and the pace of development makes it impossible to update and run conventional large reservoir models to evaluate all the possibilities. The DDRM solution is quick and appropriate, matching method to question. New technology used in this case includes model construction with uncertain objects, dynamic characterization by streamline simulation on the fly, detail created locally when required and automated, rapid model updating.